DESCRIPTION: The goal of this work is to understand the influence of chemical architecture; i.e., the constituent atoms and chemical bonds, on the conformation, properties, and interactions of nucleic acids. The research is computational, involving knowledge-based and classical (all-atom) energy potential calculations, molecular modeling, developments and applications of polymer chain statistics, and Monte Carlo simulation studies. The knowledge-based potentials incorporate observed effects of primary sequence of primary sequence on local deformations of base and backbone geometry in DNA crystals. The all-atom calculations supplement the experimentally characterized conformations and help to decipher the driving forces between different forms. The combination of modeling and polymer statistical mechanics demonstrates how variations in local conformation translate into overall changes in macromolecular structure. The latter studies provide an additional check of the knowledge-based and all-atom potentials. Their validity or rejection depends upon the correspondence of calculated with observed configuration-dependent properties. The Monte Carlo studies offer, in addition, a means to study the distributions of molecular conformations and the flexibility of the nucleic acid as a whole. Current interests include: (1) effects of base sequence and protein binding on the intrinsic structure and deformability of dimer steps and interpretations of long-range nucleic acid structure and sequence effects at a structural level; (2) predictions and analyses of high energy "activated" structures, such as severely protein-deformed or physically stretched duplexes; (3) conformational interdependence and transition pathways of base and sugar-phosphate backbone.